BackgroundThe dominantly inherited periodic fever syndromes are a group of disorders characterized by episodes of fever with serosal, synovial, conjunctival, or cutaneous inflammation; over time some patients also develop systemic AA amyloidosis. These conditions can be distinguished clinically from the recessively inherited familial Mediterranean fever (FMF) by virtue of their prolonged (greater than 1 week) attacks, periorbital edema/conjunctivitis, migratory erythema, and poor therapeutic response to colchicine. Prior to our discovery of a common genetic etiology, the dominantly inherited periodic fever syndromes subsumed several diagnostic entities, including familial Hibernian fever (FHF), benign autosomal dominant familial periodic fever, and autosomal dominant periodic fever with amyloidosis. In early 1998 a collaborator in Adelaide established genetic linkage to chromosome 12p13 for benign autosomal dominant familial periodic fever, and shortly thereafter a collaborator in London established linkage to the same region for FHF. Objective of Present StudiesDuring the 1999 reporting period our objective was to establish a positional cloning project that would ultimately identify the periodic fever gene on chromosome 12p13. Since we were able to achieve this objective in a shorter time than anticipated, we established additional objectives: a) to investigate how mutations in this gene cause disease; b) to perform mutational screening on selected referrals to the Clinical Center, and to study genotype-phenotype relationships; and c) to undertake pilot studies on the treatment of the dominantly inherited periodic fever syndromes.Results of the Last YearIdentification of the gene causing dominantly inherited periodic fevers: Pooling pedigrees from the NIH and our collaborators, we were able to define an approximately 8 cM candidate interval that included the gene (TNFRSF1A, formerly TNFR1) encoding the 55 kDa tumor necrosis factor (TNF) receptor, as well as several other plausible candidates. This TNF receptor protein is variously denoted as TNFR1, p55, and CD120a. By genomic sequencing, we identified 6 different mutations in TNFRSF1A in 7 different families with dominantly inherited periodic fevers. Five of the 6 mutations involve missense substitutions in cysteines that would directly disrupt extracellular disulfide bonds, while the sixth mutation is adjacent to one of these cysteines and may indirectly inhibit disulfide bonding. Families with mutations were not restricted to Irish or Scottish ancestry. We proposed the acronym TRAPS (TNF receptor-associated periodic syndrome) to recognize the common pathogenesis in all of these families of varied ethnic backgrounds and clinical phenotypes. The fact that TNFR1 associates into homotrimers may account for the dominant mode of inheritance, since a mutation in even one of the three chains could lead to significant alterations in the complex.Studies of the molecular pathogenesis of disease: In collaboration with the Lymphocyte Cell Biology Section of the Arthritis and Rheumatism Branch, we investigated the possible mechanisms by which these mutations might cause a hyperinflammatory phenotype. There was no increase in the binding affinity of radiolabeled TNF for leukocytes from patients as compared with controls. We also considered the possibility that disease-associated mutations might cause constitutive activation, perhaps by permitting intermolecular disulfide homodimerization and ligand-independent activation. This was tested by studying the in vitro production of IL-6, a TNF-responsive cytokine, by leukocytes from patients. Baseline IL-6 production was not increased, and the dose- response for purified mononuclear cells stimulated with TNF was similar to controls, thus arguing against constitutive activation.We did, however, find evidence that these mutations lead to a specific defect in the activation-induced shedding of the 55 kDa TNF receptor. The first suggestion that this might be the case was the observation that 27 mutation-positive individuals with 4 different mutations had about 50% of the normal level of circulating soluble p55, and a markedly blunted increase in soluble p55 during attacks (relative to patients with other inflammatory conditions). An intensive study of leukocytes from 3 patients with the C52F mutation showed increased resting levels of membrane TNFR1 relative to normal controls. Stimulation with phorbol ester (PMA) normally induces metalloprotease mediated cleavage of TNFR1 and several other receptors from the cell surface. By flow cytometry, PMA-induced clearance of TNFR1 from the surfaces of both granulocytes and monocytes was markedly reduced relative to normal controls. In contrast, clearance of the 75 kDa TNF receptor (p75) in C52F patients was comparable to controls, thus arguing against a generalized defect in receptor clearance. Moreover, when patient or control leukocytes were cultured with PMA, levels of soluble p55 in the supernatants were much lower for the C52F patients than for controls, while the levels of soluble p75 were similar in patient and control supernatants. Taken together, these data indicate that one important physiologic consequence of the C52F mutation is impaired cleavage of the TNFR1 ectodomain, thus subverting normal homeostasis by permitting repeated stimulation through the cell-bound receptor, and leading to diminished levels of potentially antagonistic circulating p55. TRAPS is the first human disease shown to be caused by a defect in the cleavage of cytokine receptors. Mutational screening and genotype-phenotype studies: Subsequent to our initial report describing the first six TNFRSF1A mutations, we have screened additional families referred for evaluation of periodic fever. Through these efforts we have identified an additional 7 mutations, all of which affect the extracellular domain of the p55 TNF receptor. Six of the 7 are missense mutations, and 2 affect cysteines involved in disulfide bond formation. The seventh mutation is a single base substitution that creates a splice acceptor site in intron 2, resulting in the addition of 4 amino acids to the extracellular domain. This mutation was initially detected by genomic sequencing, and the effect on splicing has been confirmed by cDNA sequencing of RT-PCR products. It is noteworthy that among the 16 families in which we have found TNFRSF1A mutations, there are 13 different mutations. The aforementioned splicing mutation was seen in 2 families, and the T50M mutation was observed in 3 families. However, by microsatellite haplotype analysis, even these represent independent mutational events. This is in stark contrast to FMF, where several founder mutations/haplotypes account for the majority of carrier chromosomes in several populations. It is also remarkable that, despite the fact that TRAPS is seen in a variety of ethnic groups, fully 60% of our families are of Scottish or Irish ancestry, raising the possibility of a permissive genotype at a second locus in the Irish and Scottish populations. Careful observation of families with TNFRSF1A mutations reveals a much broader clinical spectrum than had been described for the prototypic FHF. Systemic AA amyloidosis has been documented in families with the C30R, C33Y, C33G, C52F, and C88Y mutations. A rash similar to that seen in dermatomyositis has been observed in one member of a family with the C30S mutation. Consistent with the known biologic effects of TNF in promoting anorexia and cachexia, TRAPS patients often show significant weight loss during their febrile attacks, and one patient with the R92Q mutation has marked wasting and lipoatrophy. Given the expression of TNF receptors in the central nervous system, and the presence of neurologic abnormalities in TNF transgenic mice, we are actively investigating the potential for neurologic symptoms in TRAPS.Therapeutic studies: During the last 6 months we have begun a pilot study of the use of recombinant p75 TNFR:Fc fusion protein (etanercept) in the treatment of TRAPS. This agent would be expected to reduce TNF-dependent activation of inflammation, and might make up for the observed deficiency of circulating soluble p55. A total of 7 TRAPS patients have been treated with etanercept; 3 have the C52F mutation, 2 have the C33G mutation, and 1 each have the C30S and T50M mutations. The drug has been well-tolerated in all 7 patients. The clinical response in the best studied case, a patient with the C52F mutation, has been dramatic, and the preliminary experience with the remaining 6 has been favorable. Conclusions and SignificanceDuring the last year we substantially increased our understanding of the dominantly inherited periodic fever syndromes. First, we discovered that these disorders are caused by mutations in TNFRSF1A, the gene that encodes the 55 kDa receptor for TNF. We have suggested the new acronym TRAPS (TNF receptor-associated periodic syndrome) to include all mutation-positive patients, regardless of their ethnic backgrounds or the specific features of their periodic inflammation. Five of the 6 initial mutations directly disrupt intrachain disulfide bonds in the extracellular domain, and functional studies of leukocytes from patients implicate a defect in activation-induced ectodomain-cleavage. This latter observation may serve as a paradigm for additional as yet undiscovered mutations in other members of the TNF receptor family. Sequencing studies of new Clinical Center referrals have brought the total number of disease-associated mutations to 13, and careful clinical observations of these families are likely to deepen our understanding of the role of TNF pathways in human disease. Our data indicate that TNF receptor mutations are not restricted to the Irish and Scottish populations, though there is an unexplained enrichment for these ethnicities among affected families. During the next year, our objectives will be:1) continuing mutational and genotype-phenotype studies of selected patients and families referred to the NIH Clinical Center; 2) developing a blinded, placebo-controlled crossover study to evaluate the efficacy of etanercept in TRAPS more systematically; 3) collaborative studies with the Lymphocyte Cell Biology Section of ARB to study signaling in cells transfected with TNFR1 bearing disease- associated mutations; and 4) developing a TRAPS knockin mouse.